Nitrous oxide, N.sub.2 O, frequently referred to as laughing gas, recently has been increasingly investigated as an undesirable component of gaseous emissions. Although formed in nature from bacterial action in soils and oceans, the levels associated with such "natural" emissions are not an environmental concern, a concern which arises in part from its relative unreactivity contributing to ozone layer depletion, and in part from its acting as a greenhouse gas, analogous in effect to carbon dioxide but far more potent. However, atmospheric levels of N.sub.2 O are found to be increasing bringing with it increased concern for its accumulation. One small but significant source of N.sub.2 O is that of adipic acid manufacture, where N.sub.2 O formation accompanies the nitric acid oxidation of cyclohexanone. Exit gases from the aforementioned reaction may contain quite high concentrations (about 30 volume percent) of N.sub.2 O and are discharged directly into the atmosphere. Off gases from nitric acid plants also are being recognized as a source of N.sub.2 O, although perhaps in significantly lower exit gas concentration, but nonetheless in substantial total amounts. Other N.sub.2 O-emission sources are the manufacture of hydroxylamine derivatives, caprolactam production, and the low temperature combustion of nitrogen-containing materials, as in fluid bed incinerators. With increasing global environmental sensitivity generally, and with more stringent local requirements particularly, there has arisen a need for implementing a process for N.sub.2 O destruction which addresses the concerns attached to emissions containing substantial amounts and/or concentrations of N.sub.2 O.
Several catalyst systems are known to effect the thermal decomposition of nitrous oxide to nitrogen and oxygen. However effective these may be, some inherent characteristics of nitrous oxide decomposition largely independent of the particular catalyst system used introduce complexities which the process of this invention addresses. In particular, even though N.sub.2 O decomposition is a highly exothermic reaction, whose heat of reaction is approximately 19.5 kcal mole.sup.-1, its decomposition is initiated by most catalytic systems at a temperature of several hundred degrees centigrade. Where the catalyst is used as a solid bed, the large reaction exotherm coupled with a high space velocity of the N.sub.2 O-containing gas stream through the reactor means that there is a considerable temperature increase within and along the catalyst bed, an increase which easily can reach several hundred degrees centigrade. But such a large temperature increase across the catalyst bed can have several adverse consequences. One potential detriment is sintering of the catalyst and/or catalyst support leading to a reduction of catalyst activity and a reduction in catalyst life. Another detriment is that many catalysts used for the decomposition of N.sub.2 O also can effect the reaction of nitrogen and oxygen to form NO.sub.x. Thus, it is often desirable to have some maximum temperature, T.sub.max, in the catalyst bed to avoid the foregoing detriments. This maximum temperature will depend primarily on the catalyst and should not be exceeded despite changes in the N.sub.2 O-containing feed, feed rate, the temperature necessary to initiate N.sub.2 O decomposition, and other reaction variables.
Although many ways are possible to ensure this outcome, we have found that a particularly effective means is to cool some portion of the effluent gas and to recycle the cooled effluent gas to one or more points along the decomposition zone. By varying the amount of effluent gas recycled, the degree to which it is cooled, the points at which it is recycled to the decomposition zone, and/or feed flow rate, it is possible to maintain the temperature everywhere in the decomposition zone less than T.sub.max while maintaining a high throughput of the N.sub.2 O-containing waste gases in the decomposition zone.